Free RAID-AI Framework Helps Organisations Check Their AI Systems Against Ransomware

A new free tool is available to help organisations find and fix weaknesses that could leave their AI systems open to ransomware. The Ransomware Assessment and Incident Defense Framework for AI Systems (RAID-AI) was developed over two years by Behnaz Karimi, co-lead of the OWASP AI Bill of Materials initiative, with help from fellow co-lead Yuvaraj Govindarajulu. It covers ransomware attacks on AI systems whether or not the attacker uses AI.

Karimi told Infosecurity Magazine that ransomware attacks targeting AI systems are slowly starting to appear. She pointed to an April 2025 report describing ransomware actors using data poisoning techniques, and to an alleged attack on pharmaceutical company Novo Nordisk, where hackers claimed to have stolen AI models.

RAID-AI is a hands-on web console. Teams rate their own controls manually on a scale from zero to five, rather than connecting or scanning their systems. Controls are split into general areas, such as recovery, training and sensitive-data protection, and AI-specific areas covering data, tools and runtime. The results map to NIST Cybersecurity Framework 2.0 and CISA Cybersecurity Performance Goals 2.0. Users can also map adversarial attacks using MITRE ATT&CK, MITRE ATLAS and NIST Adversarial.